So, after designing my own, sort of curious how to make it more temperature stable. These guys did it right it seems:
There exist a second way to do this, in that case you drive the coils into saturation halfway through the period, and 'tune' the pick up coils to the second harmonic. Somehow this seems to cancel temperature influences. In my fluxgate heating it up can make several degrees difference.
Now I see these guys drive with 15625 Hz, and I see 3 big pot cores ... could be a LC filter at 31250 Hz...
Also 3 separate 1 axis magnetometers make it much more easy to align. There is also some other patent on this, but I tried downloading the pdf from freepatentsonline.com without any luck, butdescribes some .. art, no pictures.
It looks as if I was right about the Cassini magnetometer using second harmonic, I found a Themis magnetometer paper free for download:it has a nice block diagram, is FPGA based, and uses the second harmonic method, uses a feedback coil.
Enough new material for more experiments. I want to see that earth magnetic field changing too:
I am going to log the values with the current mag_pic tonight during the night, as then (hopefully) no magnets will move about in the house, this is a big problem, it picks up things very far away. Will just write a script and get status every minute or so, and parse it later.